Wasn’t there a point in the Atlanta/New England game earlier this year where the Pats went for it on like 4th and 2 on their own 30 when they didn’t have to, and converted it, and drove down to win the game? I wonder where the pages and pages out outrage were then.
There was a study about a year ago that concluded that football coaches go for it on 4th down far less often than is optimal-as in, if they did follow said optimal strategy, they likely would have c. 60% fewer punts and field goal attempts. I can’t find it in Google, but memory serves we talked about it here-just came up empty on my first 90 second search tho. Did find an article on a high school team which never punts tho.
That’s beside the point, though. Conversions inside the 10 are harder because the defense doesn’t have to worry about someone getting behind them. The Colts did - the yardage wouldn’t have made any difference, but their coverage still had to account for a deep pass.
I’m not arguing that the decision wasn’t questionable, in the sense that a case could be made either way. I am arguing that Belichick made the correct call.
Stripping all the numbers and stuff out of it, I think the decision boils down to this: is Tom Brady more likely to get you two yards than Peyton Manning is to drive down the field? I think, regardless of any other consideration, that Brady is at least as likely to pick up 2 yards as Manning is to drive 65.
Yes, it was late third quarter, the Pats were at home, 4th and 1 from their own 24, up 16 - 10 and it was against the Falcons. Not arguing, just providing context.
Actually, you don’t. You were fine with your analysis of the possibilities of going for it, and reasonably fine with your probable punt.
Then you totally failed to give the relevant data for the chance of the Colts driving the required distance in the required time, given the number of timeouts they had.
Hell, screw the time and timeout factors. Provide the probabilities of the Colts driving the required distance for a touchdown. That is, how many drives starting from roughly that part of the field do the Colts convert into touchdowns compared to all possible drives started from that distance?
My guess is the number is much closer to that 20% than you are willing to admit, which is why you stopped providing the data.
Further, I take issue with the data you provided. For example, asserting that the generic data on a generic 4th and 2 is applicable here is silly. This isn’t a 4th and 2 in the middle of the game when nothing rides on it. Now, if you were to quote data on 4th and 2 where the game is on the line (say 4th and 2 in the last 3 minutes of the game), that might be better. And for all we know, that might actually be a HIGHER success rate, though I tend to doubt it. As a rule, defenses tend to step up in such situations; the offense is ALWAYS jazzed up to try and make a short 4th down situation.
But, more importantly, the chances of the Colts scoring a TD based on the starting yardage when the game is on the line is likely much higher than their chance of scoring a TD from that yardage in a generic situation. And it doesn’t take much change in the numbers to make the decision look foolish.
For example: Let’s say the chance of conversion is 50%. Let’s say that the chance of success by the Colts on a blow conversion is 80%.
Now, it’s a 40% chance the Colts score the winner (plus whatever chance they have of gaining the ball with about 30 sec. left and then scoring a TD, admittedly relatively de minimis). Now compare that to the chance of the Colts scoring a TD with 2 min. left and going 70+ yards to do it.
And that’s how I look at it, and how I looked at it at the time. Obviously, Belichik thought differently. Clearly, he was of the opinion either that the chance of conversion was higher, or that the chances of the Patriot’s D stopping Manning and the Colts after a punt was relatively low.
All of this points out the real reason it was probably a bad decision: there simply is no way that you can calculate these probabilities out so finely. To do so eventually requires pulling something out of your ass. The risk acceptant person will run the risk that he’s made a bad calculation; the risk averse person will not. When the probability is relatively high that your failure will result in a loss, you probably should tend to estimate things conservatively. Belichik didn’t, and got what he deserved for running the risk.
It is more than that, since it isn’t 100% certain the Colts score from the 28. So the question would be is it more likely Brady gets 2 yards or Manning would have scored from the Pat 28, but not his own 35.
If you want to play at home
http://belichick-decision.heroku.com/
I only just saw the 4th down play…was still mourning the Steelers’ egg-laying against the Bengals to watch Colts-Pats.
Anyway, I haven’t heard a word said about this: was Brady’s pass intended for Faulk at all? From one of the replay angles you could see that Moss was in the line of the pass, and it looked like Faulk reached up and tried to grab a pass that wasn’t intended for him. The way Moss reacted to the play, it was almost as if he was getting ready to break towards the ball.
All in all, it seemed like a terribly-designed play.
This works against you more than for you. Unless you want to play the “it’s almost guaranteed that the juggernaut Colts offense score from 35 out, but not that likely if they get the ball from 70 yards out.” game.
Yep. It’s unfortunate that Belichick’s certain mistakes are being overlooked because of his possible mistake- but the truth is that was a terrible play-call.
If they wanted to throw to Faulk that’s fine, but at least send him out on a swing pass or something so the safety won’t be lined up three inches from him when you throw the pass. The Colts’ pass-rush wasn’t getting home, so Brady should have had time to actually make reads instead of automatically going to Faulk.
Again, the Colts had just done that - 79 yards in 1:49 without taking a timeout. Having seen that happen and knowing that a) your defense is exhausted, and b) you won’t be able to sub players against Indy’s no-huddle has to play a role in the decision you’re making.
Not to mention that they were down to two good pass-rushers, with Tully Banta-Cain out (and Adalius Thomas was supposedly playing with some kind of foot injury).
Totally unrelated question: what are the odds that Matt Light is shown the door at the end of the season? That Sebastian Bullhorn (or whatever his name is - Verhoven?) totally shut down Dwight Freeney.
No, it didn’t. Watch the replay. They rushed 5 (and had Brady under very heavy pressure), dropped 1 LB into a short middle zone, and the other 5 were all in tight man coverage. Nobody was more than 5 yards past the LOS.
Interestingly, they DID have a man semi-deep until they motioned Faulk into the slot (the deep S came up). When they went empty backfield, they helped the Colts out, IMO.
Right - see above, where I pointed out that the play called was a bad one.
There seems to be a trend this season for teams to line up with an empty backfield on 3rd/4th and short, which I don’t understand at all because it allows the defense to totally ignore the run, and play action. The Steelers have been particularly egregious in this regard.
One other thing, and I better leave it at one thing because I could write about this all day, being a number-cruncher myself.
It’s all well and good to say “the Patriots historically have a 63.2% success rate in 4th-and-2 plays.” But that figure is an average based on a wide range of data points. Most notably, it doesn’t take into account situational factors. And a big situational factor would have to be “on the road, in one of the loudest stadiums in the league.” (As an aside, one of the reasons the play call was so bad involved exactly that: Faulk and Moss were too far away from center to be able to hear an audible.) You’d have to think that the Patriots’ chance of success would have to be significantly lower than average in this situation. Would that be enough to make the play call a bad choice? I don’t know.
If you completely isolate the fourth-down call, it was potentially a good one. But it was in the context of a bad call. The Patriots should have gone into third-and-two in a two-down mindset. They should have been running on third, or maybe setting up a screen pass. Fourth-and-two with the clock stopped shouldn’t have happened at all. The way I look at the fourth-down call is like the guy who stops for directions only to be told, “If I were you, I wouldn’t have started from here.”
I love stats geekery in sports, but the difference between baseball and football is that decisions can’t be made in isolation. You can make an objective decision on pitcher-batter matchups because there are really only two people involved. It’s a lot harder to make those kinds of analyses in football.
I agree; but the point was that contra your declaration, the Colts did not defend the deep pass.
Yes, but that doesn’t mean they shouldn’t. With the field wide open a swing pass to a guy out of the backfield, slant, or crossing route would have been very effective. The fact that the Pats chose not to send anyone on those routes does not change whether or not the decision to go for it itself was a good one. As you note, when the Pats did have Faulk lined up in the backfield the Colts were at least defending the middle.
An average team will score from approximately 60 yards out with one timeout approximately 30% of the time. This is Peyton Fricking Manning, so I adjusted upward by 10% to reach my estimate of 40%.
But… but… you are inflating the chances of the Colts scoring from the 28 (from the historical 63% to 80%), but refusing to inflate the chances of the Colts scoring from their own 37 by a like margin. If it’s more likely that they’ll score when the game is on the line, then it’s more likely that they’ll score when the game is on the line from 63 yards away as well. The historical average for 63 yards away at the 2:00-minute warning with one timeout is 30%. Adding Peyton Manning, Reggie Wayne, et al, a gassed defense, the fact that the Colts were at home, and the end-of-game situation, I don’t see how you can’t inflate that 30% to at least 40% - which means even using your numbers for the other figurings (which inflate the chances of the Colts scoring by a full 17% over average from the 28), it’s a wash. So if you assume that:
- The Patriots, in spite of having an elite offense, will convert on fourth and 2 less often than an average team will throughout history; and
- The Colts will convert at 17% better than average starting from the 28; and
- The Colts will convert at only 10% better than average starting from their own 37:
Even then, you’ve still only shown that Belicheck’s decision was a coin flip.
In the case in question, the “failure would have resulted in a loss” no matter what choice was made. If Belicheck decides to punt and that gambit fails - ie, Manning brings his team down and scores anyway - that’s still a loss.
Sigh. They still defended the middle; watch the damn replay.
Please try to follow the conversation.
–You said they had a higher than 60% chance of succeeding on 4th and 2.
–I said it was lower, more like 50%, given that it was more akin to a goal line situation, with no need to defend deep.
–You said, at 1:06, that they did need to defend deep. You even put in italics.
–I pointed to the replay, which shows the opposite. No deep defense.
–You changed the subject to the specific play call.
–My point stands – fourth and two, in an effectivelly goal-line situation, is different than 4th and 2 generally.